* OPEN STATA OUTPUT FILE LOG *

log using "C:\Users\jungy\Dropbox\DISCRIMINATION PROJECT\Organizational Diversity\Statistics\Krause & Park.Authority Differentials.APPENDIX I RESULTS.08-07-2024.smcl", replace  

  


  
  
***** SUPPLEMENTARY APPENDIX STATISTICAL ANALYSES:  APPENDIX I: CONSIDERING SPILLOVER OR CONTAGION EFFECTS ACROSS SOCIAL IDENTITY 'OUT-GROUPS' /// ******



*** ACCESS DATABASE FOR THE PROJECT: FEVS DATA FROM 2010-2019 AND 'MATCHED' OPM DATA: POST-ESTIMATION OF MANUSCRIPT RESULTS ****


use "C:\Users\jungy\Dropbox\DISCRIMINATION PROJECT\Organizational Diversity\Statistics\2010-2019_DATA FINAL.08-07-2024.post-estimation.dta", replace 

   

 
**************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************



*** MODEL I1: Women SGPD contagion for Minority/Non-Minority Respondents as an Added Set of Covariates ***


regress lndiversity2zeroadj  c.ln_ratio_fmsup_fmsub##i.gender   ln_ratio_fem_tot_men_tot   minority  supervisor  topoffgender_2 lntotworkforce_count  ln_professionals_total_ratio i.agencyid i.year c.ln_ratio_fmsup_fmsub##i.minority, vce(cluster agencyid)
*
estat ic
*
** BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN GENDERED RESPONDENTS **

lincom c.ln_ratio_fmsup_fmsub
*
*
lincom  1.gender#c.ln_ratio_fmsup_fmsub
*
*
*
*
*
*
*
*

*** MODEL I2: Minority SGPD contagion for Women/Men Respondents as an Added Set of Covariates  ***

** BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN MINORITY/NON-MINORITY RESPONDENTS **

regress  lndiversity2zeroadj  c.ln_ratio_mnmsup_mnmsub##i.minority   ln_ratio_min_tot_nmin_tot  gender supervisor  topoffminority_2 lntotworkforce_count  ln_professionals_total_ratio   i.agencyid i.year c.ln_ratio_mnmsup_mnmsub##i.gender, vce(cluster agencyid)
*
estat ic
*


lincom c.ln_ratio_mnmsup_mnmsub
*
*
lincom  1.minority#c.ln_ratio_mnmsup_mnmsub
*
*
*
*



*** MODEL I3: HETEROGENEOUS Women SGPD contagion for HETEROGENEOUS Minority/Non-Minority RESPONDENTS as an Added Set of Covariates  ***

regress lndiversity2zeroadj  c.ln_ratio_fmsup_fmsub##i.women_het   ln_ratio_fem_tot_men_tot  minority  supervisor  topoffgender_2 lntotworkforce_count  ln_professionals_total_ratio  i.agencyid i.year  c.ln_ratio_fmsup_fmsub##i.minority_het, vce(cluster agencyid)
*
estat ic
*
** BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN GENDERED RESPONDENTS **

lincom c.ln_ratio_fmsup_fmsub
*
*
lincom  1.women_het#c.ln_ratio_fmsup_fmsub
*
*
lincom  2.women_het#c.ln_ratio_fmsup_fmsub
*
*
lincom  2.women_het#c.ln_ratio_fmsup_fmsub -  1.women_het#c.ln_ratio_fmsup_fmsub
*
*
*
*



*** MODEL I4: HETEROGENEOUS Minority SGPD contagion for Minority/Non-Minority Respondents for HETEROGENEOUS Minority/Non-Minority RESPONDENTS as an Added Set of Covariates  ***

regress lndiversity2zeroadj  c.ln_ratio_mnmsup_mnmsub##i.minority_het   ln_ratio_min_tot_nmin_tot   gender supervisor  topoffminority_2 lntotworkforce_count  ln_professionals_total_ratio  i.agencyid i.year  c.ln_ratio_mnmsup_mnmsub##i.women_het, vce(cluster agencyid)
*
estat ic
*
** BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN GENDERED RESPONDENTS **

** BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN MINORITY/NON-MINORITY RESPONDENTS **

lincom c.ln_ratio_mnmsup_mnmsub
*
*
lincom  1.minority_het#c.ln_ratio_mnmsup_mnmsub
*
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub
*
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub -  1.minority_het#c.ln_ratio_mnmsup_mnmsub
*
*
*
*


**************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************
*********************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************


*********** estimating the above models making a more granular distinction between Non-Supervisory and Supervisory Respondents through a triple interaction term ***********


********************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************



*** MODEL I5: Women Absolute SGPD contagion for Minority/Non-Minority Respondents as an Added Set of Covariates  -- CONTROLLING FOR RACIAL/ETHNIC MINORITY SUPERVISORY EMPLOYEE IDENTITY GROUP DIFFERENTIAL ***

regress  lndiversity2zeroadj  c.ln_ratio_fmsup_fmsub##i.gender##i.supervisor   ln_ratio_fem_tot_men_tot   minority  topoffgender_2 lntotworkforce_count  ln_professionals_total_ratio   i.agencyid i.year c.ln_ratio_fmsup_fmsub##i.minority##i.supervisor , vce(cluster agencyid)
*
estat ic
*
** BY NON-SUPERVISORS RESPONDENT: WITHIN-IDENTITY "OUT" GROUP STATUS DIFFERENTIAL BETWEEN GENDERED RESPONDENTS **

lincom c.ln_ratio_fmsup_fmsub 
*
*
lincom  1.gender#c.ln_ratio_fmsup_fmsub

*
*
*
*

** BY SUPERVISOR RESPONDENT: WITHIN-IDENTITY "OUT" GROUP STATUS DIFFERENTIAL BETWEEN GENDERED RESPONDENTS **

lincom c.ln_ratio_fmsup_fmsub + 1.supervisor#c.ln_ratio_fmsup_fmsub
*
*
lincom  1.gender#c.ln_ratio_fmsup_fmsub +  1.gender#1.supervisor#c.ln_ratio_fmsup_fmsub








*** MODEL I6: Minority Absolute SGPD contagion for Women/Men Respondents as an Added Set of Covariates  -- CONTROLLING FOR GENDER SUPERVISORY EMPLOYEE IDENTITY GROUP DIFFERENTIAL ***


regress lndiversity2zeroadj  c.ln_ratio_mnmsup_mnmsub##i.minority##i.supervisor  ln_ratio_min_tot_nmin_tot   gender  topoffminority_2 lntotworkforce_count  ln_professionals_total_ratio   i.agencyid i.year c.ln_ratio_mnmsup_mnmsub##i.gender##i.supervisor, vce(cluster agencyid)
*
estat ic
*

** BY NON-SUPERVISORS RESPONDENT: WITHIN-IDENTITY "OUT" GROUP STATUS DIFFERENTIAL BETWEEN MINORITY/NON-MINORITY RESPONDENTS **

lincom c.ln_ratio_mnmsup_mnmsub
*
*
lincom  1.minority#c.ln_ratio_mnmsup_mnmsub

*
*
*
*

** BY SUPERVISOR RESPONDENT: WITHIN-IDENTITY "OUT" GROUP STATUS DIFFERENTIAL BETWEEN MINORITY/NON-MINORITY RESPONDENTS **

lincom c.ln_ratio_mnmsup_mnmsub + 1.supervisor#c.ln_ratio_mnmsup_mnmsub
*
*
lincom  1.minority#c.ln_ratio_mnmsup_mnmsub +  1.minority#1.supervisor#c.ln_ratio_mnmsup_mnmsub
*
*
*
* SUPERVISOR - NON-SUPERVISORY DIFFERENCE AMONG MINORITY RESPONDENT DIFFERENCES 
 
lincom  1.minority#c.ln_ratio_mnmsup_mnmsub +  1.minority#1.supervisor#c.ln_ratio_mnmsup_mnmsub - (1.minority#c.ln_ratio_mnmsup_mnmsub)




*****************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************



   
*** MODEL I7: Women Relative SGPD contagion for Minority/Non-Minority Respondents as an Added Set of Covariates -- HETEROGENOUS WOMEN & MINORITY REPSONDENT EFFECTS ***


regress lndiversity2zeroadj  c.ln_ratio_fmsup_fmsub##i.women_het##i.supervisor   ln_ratio_fem_tot_men_tot   minority   topoffgender_2 lntotworkforce_count  ln_professionals_total_ratio   i.agencyid i.year c.ln_ratio_fmsup_fmsub##i.minority_het##i.supervisor, vce(cluster agencyid)
*
estat ic
*
** BY NON-SUPERVISORS RESPONDENT: BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN GENDERED RESPONDENTS **

lincom c.ln_ratio_fmsup_fmsub
*
*
lincom  1.women_het#c.ln_ratio_fmsup_fmsub
*
*
lincom  2.women_het#c.ln_ratio_fmsup_fmsub
*
*
*
lincom 2.women_het#c.ln_ratio_fmsup_fmsub -  1.women_het#c.ln_ratio_fmsup_fmsub
*
*
*
*


** BY SUPERVISOR RESPONDENT:BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEENGENDERED RESPONDENTS **

lincom c.ln_ratio_fmsup_fmsub + 1.supervisor#c.ln_ratio_fmsup_fmsub
*
*
lincom  1.women_het#c.ln_ratio_fmsup_fmsub +  1.women_het#1.supervisor#c.ln_ratio_fmsup_fmsub
*
*
lincom  2.women_het#c.ln_ratio_fmsup_fmsub +  2.women_het#1.supervisor#c.ln_ratio_fmsup_fmsub
*
*
*
lincom  2.women_het#c.ln_ratio_fmsup_fmsub +  2.women_het#1.supervisor#c.ln_ratio_fmsup_fmsub - (1.women_het#c.ln_ratio_fmsup_fmsub +  1.women_het#1.supervisor#c.ln_ratio_fmsup_fmsub)



* SUPERVISOR - NON-SUPERVISORY DIFFERENCE AMONG WOMEN RESPONDENT DIFFERENCES [NON-MINORITY WOMEN RESPONDENTS FOLLOWED BY MINORITY WOMEN RESPONDENTS] -- DO NOT PLOT IN GRAPHS [ONLY FOR TEXT]!

lincom  1.women_het#c.ln_ratio_fmsup_fmsub +  1.women_het#1.supervisor#c.ln_ratio_fmsup_fmsub  - (1.women_het#c.ln_ratio_fmsup_fmsub)
*
lincom  2.women_het#c.ln_ratio_fmsup_fmsub +  2.women_het#1.supervisor#c.ln_ratio_fmsup_fmsub  - (2.women_het#c.ln_ratio_fmsup_fmsub)










   
*** MODEL I8: Minority Relative SGPD contagion for Women/Men Respondents as an Added Set of Covariates -- HETEROGENOUS WOMEN & MINORITY REPSONDENT EFFECTS  ***

regress  lndiversity2zeroadj  c.ln_ratio_mnmsup_mnmsub##i.minority_het##i.supervisor  ln_ratio_min_tot_nmin_tot   gender  topoffminority_2 lntotworkforce_count  ln_professionals_total_ratio  i.agencyid i.year c.ln_ratio_mnmsup_mnmsub##i.gender##i.supervisor  if e(sample), vce(cluster agencyid)
*
estat ic
*

** BY NON-SUPERVISORS RESPONDENT: BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN MINORITY/NON-MINORITY RESPONDENTS **

lincom c.ln_ratio_mnmsup_mnmsub
*
*
lincom  1.minority_het#c.ln_ratio_mnmsup_mnmsub
*
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub
*
*
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub - 1.minority_het#c.ln_ratio_mnmsup_mnmsub

*
*
*
*
*


** BY SUPERVISOR RESPONDENT: BETWEEN-IDENTITY "IN" GROUP VERSUS "OUT" GROUP STATUS DIFFERENTIAL BETWEEN MINORITY/NON-MINORITY RESPONDENTS **

lincom c.ln_ratio_mnmsup_mnmsub+ 1.supervisor#c.ln_ratio_mnmsup_mnmsub
*
*
lincom  1.minority_het#c.ln_ratio_mnmsup_mnmsub +  1.minority_het#1.supervisor#c.ln_ratio_mnmsup_mnmsub
*
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub +  2.minority_het#1.supervisor#c.ln_ratio_mnmsup_mnmsub
*
*
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub +  2.minority_het#1.supervisor#c.ln_ratio_mnmsup_mnmsub - (1.minority_het#c.ln_ratio_mnmsup_mnmsub +  1.minority_het#1.supervisor#c.ln_ratio_mnmsup_mnmsub)
*
*
*
*


* SUPERVISOR - NON-SUPERVISORY DIFFERENCE AMONG MINORITY RESPONDENT DIFFERENCES [MINORITY MEN RESPONDENTS FOLLOWED BY MINORITY WOMEN RESPONDENTS] -- DO NOT PLOT IN GRAPHS [ONLY FOR TEXT]!

lincom  1.minority_het#c.ln_ratio_mnmsup_mnmsub +  1.minority_het#1.supervisor#c.ln_ratio_mnmsup_mnmsub - (1.minority_het#c.ln_ratio_mnmsup_mnmsub)
*
lincom  2.minority_het#c.ln_ratio_mnmsup_mnmsub +  2.minority_het#1.supervisor#c.ln_ratio_mnmsup_mnmsub - (2.minority_het#c.ln_ratio_mnmsup_mnmsub)


clear

*** FIGURE I1: PLOT ELASTICITY MARGINAL EFFECTS FOR MODELS I1 & I2 USING LINCOMS ABOVE FOR EACH MODEL: PATTERN AFTER COMPARABLE SET OF MANUSCRIPT GRAPHICS/FIGURES [FIGURE 1] **
import excel "C:\Users\jungy\Dropbox\DISCRIMINATION PROJECT\Organizational Diversity\Statistics\figurei1.xlsx", sheet("Sheet1") firstrow
destring, replace

set scheme sj, permanently
graph set window fontface "Century Schoolbook"

twoway (rcap low95 high95 row, vert) (scatter estimates row if group ==1, msymbol(square) mcolor(orange))(scatter estimates row if group ==2, msymbol(circle_hollow) mcolor(orange))(scatter estimates row if group ==3, msymbol(square) mcolor(navy))(scatter estimates row if group ==4, msymbol(circle_hollow) mcolor(navy)), legend(row(1) order(2 "Gender" 4 "Race/Ethnicity") pos(6)) title("FIGURE I1" "Relationship Between Authority Differentials and D&I Employee Evaluations" "(By Respondent Single Social Identity Group)" "[Sensitivity of AD Estimates When Controlling for Social Identity Group Spillover Effects]", size(small)) ylabel(-0.1(.1)0.3, labsize (small) angle(horizon)) xtitle("AD Estimates [Differentials by Responsdent Single Social Identity Group]" "(Models 1&2)", size(small)) xlabel("", noticks) yline(0, lpattern(dash) lcolor(gs8)) aspect(.5)

clear

*** FIGURE I2: PLOT ELASTICITY MARGINAL EFFECTS FOR MODELS I3 & I4 USING LINCOMS ABOVE FOR EACH MODEL: PATTERN AFTER COMPARABLE SET OF MANUSCRIPT GRAPHICS/FIGURES [FIGURE 2]  **
import excel "C:\Users\jungy\Dropbox\DISCRIMINATION PROJECT\Organizational Diversity\Statistics\figurei2.xlsx", sheet("Sheet1") firstrow
destring, replace

set scheme sj, permanently
graph set window fontface "Century Schoolbook"

twoway (rcap low95 high95 row, vert) (scatter estimates row if group ==1, msymbol(square) mcolor(orange))(scatter estimates row if group ==2, msymbol(circle_hollow) mcolor(orange)) (scatter estimates row if group ==3, msymbol(diamond_hollow) mcolor(orange))(scatter estimates row if group ==4, msymbol(triangle_hollow) mcolor(orange))(scatter estimates row if group ==5, msymbol(square) mcolor(navy))(scatter estimates row if group ==6, msymbol(circle_hollow) mcolor(navy))(scatter estimates row if group ==7, msymbol(diamond_hollow) mcolor(navy))(scatter estimates row if group ==8, msymbol(triangle_hollow) mcolor(navy)), legend(row(1) order(2 "Gender" 6 "Race/Ethnicity") pos(6)) title("FIGURE I2"  "Relationship Between Authority Differentials and D&I Employee Evaluations" `"(By Respondent Intersectional Social Identity Group)"' "[Sensitivity of AD Estimates When Controlling for Social Identity Group Spillover Effects]", size(small)) ylabel(-0.1(.1)0.3, labsize (small) angle(horizon)) xtitle("Gender AD Effects: by Respondent Intersectionality Group   Race/Ethnicity AD Effects: by Respondent Intersectionality Group", size(vsmall)) xlabel("", noticks) yline(0, lpattern(dash) lcolor(gs8)) aspect(.5)
  
clear

*** FIGURE I3: PLOT ELASTICITY MARGINAL EFFECTS FOR MODELS I5-I8 USING LINCOMS ABOVE FOR EACH MODEL: PATTERN AFTER COMPARABLE SET OF MANUSCRIPT GRAPHICS/FIGURES [FIGURE 3] **
import excel "C:\Users\jungy\Dropbox\DISCRIMINATION PROJECT\Organizational Diversity\Statistics\figurei3.xlsx", sheet("Sheet1") firstrow
destring, replace

set scheme sj, permanently
graph set window fontface "Century Schoolbook"

twoway (rcap low95 high95 row, vert) (scatter estimates row if group ==1, msymbol(square) mcolor(orange))(scatter estimates row if group ==2, msymbol(square_hollow) mcolor(orange)) (scatter estimates row if group ==3, msymbol(square) mcolor(navy))(scatter estimates row if group ==4, msymbol(square_hollow) mcolor(navy))(scatter estimates row if group ==5, msymbol(square) mcolor(orange))(scatter estimates row if group ==6, msymbol(circle_hollow) mcolor(orange))(scatter estimates row if group ==7, msymbol(diamond_hollow) mcolor(orange))(scatter estimates row if group ==8, msymbol(triangle_hollow) mcolor(orange))(scatter estimates row if group ==9, msymbol(square) mcolor(navy))(scatter estimates row if group ==10, msymbol(circle_hollow) mcolor(navy))(scatter estimates row if group ==11, msymbol(diamond_hollow) mcolor(navy))(scatter estimates row if group ==12, msymbol(triangle_hollow) mcolor(navy)), legend(row(1) order(2 "Gender" 4 "Race/Ethnicity") pos(6)) title("FIGURE I3" "Relationship Between Authority Differentials and D&I Employee Evaluations" "(Non-Supervisory Respondents: Single and Intersectional Social Identity Groups)" "[Sensitivity of AD Estimates When Controlling for Social Identity Group Spillover Effects]", size(small)) ylabel(-0.05(.05)0.2, labsize (small) angle(horizon)) xtitle("AD Effects: by Respondent Single Identity Group     AD Effects: by Respondent Intersectionality Group", size(vsmall)) xlabel("", noticks) yline(0, lpattern(dash) lcolor(gs8)) aspect(.5)

clear

*** FIGURE I4: PLOT ELASTICITY MARGINAL EFFECTS FOR MODELS I5-I8 USING LINCOMS ABOVE FOR EACH MODEL: PATTERN AFTER COMPARABLE SET OF MANUSCRIPT GRAPHICS/FIGURES [FIGURE 4]  **
import excel "C:\Users\jungy\Dropbox\DISCRIMINATION PROJECT\Organizational Diversity\Statistics\figurei4.xlsx", sheet("Sheet1") firstrow
destring, replace

set scheme sj, permanently
graph set window fontface "Century Schoolbook"

twoway (rcap low95 high95 row, vert) (scatter estimates row if group ==1, msymbol(square) mcolor(orange))(scatter estimates row if group ==2, msymbol(square_hollow) mcolor(orange)) (scatter estimates row if group ==3, msymbol(square) mcolor(navy))(scatter estimates row if group ==4, msymbol(square_hollow) mcolor(navy))(scatter estimates row if group ==5, msymbol(square) mcolor(orange))(scatter estimates row if group ==6, msymbol(circle_hollow) mcolor(orange))(scatter estimates row if group ==7, msymbol(diamond_hollow) mcolor(orange))(scatter estimates row if group ==8, msymbol(triangle_hollow) mcolor(orange))(scatter estimates row if group ==9, msymbol(square) mcolor(navy))(scatter estimates row if group ==10, msymbol(circle_hollow) mcolor(navy))(scatter estimates row if group ==11, msymbol(diamond_hollow) mcolor(navy))(scatter estimates row if group ==12, msymbol(triangle_hollow) mcolor(navy)), legend(row(1) order(2 "Gender" 4 "Race/Ethnicity") pos(6)) title("FIGURE I4" "Relationship Between Authority Differentials and D&I Employee Evaluations" "(Supervisor Respondents: Single and Intersectional Social Identity Groups)" "[Sensitivity of AD Estimates When Controlling for Social Identity Group Spillover Effects]", size(small)) ylabel(-0.1(.1)0.3, labsize (small) angle(horizon)) xtitle("AD Effects: by Respondent Single Identity Group     AD Effects: by Respondent Intersectionality Group", size(vsmall)) xlabel("", noticks) yline(0, lpattern(dash) lcolor(gs8)) aspect(.5)
   

******************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************************



log close




   
   